Samples of preparations contaminated by diethylene glycol (DEG), diethylene glycol raw materials and laboratory prepared solutions were measured to get NIR spectra. Then the identi fication models were developed using the collected spectra and the spectra of distilled water, propylene glycol and the preparations without diethylene glycol. Besides, the quantification model was also established for determining the concentration of diethylene glycol in the preparations. Validation results show the identification and quantification models have ideal prediction performance. The emergency NIR models are rapid, easy to use and accurate, and can be implemented for identifying diethylene glycol raw material, screening the preparations contaminated by diethylene glycol in the markets and analyzing the concentrations of DEG.